Saliva maintains oral health. Building on our past studies of saliva formation and its alteration during pathology, we developed novel approaches to treat salivary dysfunction using principles of gene therapy. Our studies during the past year primarily addressed fundamental questions necessary to move gene therapy into the clinic for phase-I trials in two areas: irradiation (IR)-induced salivary hypofunction and systemic single protein deficiency disorders (SSPDDs). [unreadable] [unreadable] Treatment of most head and neck cancer patients includes IR. Salivary glands in the IR field suffer irreversible damage. We have tried to develop novel ways either to prevent such damage, or to restore function to damaged glands. Based on previous pre-clinical studies with a serotype 5 adenoviral (rAd5) vector encoding human aquaporin-1 (AdhAQP1) conducted in rats and miniature pigs, and an extensive safety study of AdhAQP1 in rats, we submitted a clinical protocol for testing AdhAQP1 in patients that received all required approvals, including FDA approval. In the past year, we established the infrastructure for conducting this protocol, e.g., developing case report forms and a database. Also, the clinical grade (GMP) AdhAQP1 vector was received and schemes developed for its handling and dilution with the Clinical Center Pharmaceutical Development Service. The vector specific activity was higher than anticipated, so administered doses can be decreased to one-third those originally planned, i.e., now from 4.8 x 10e7 particle units (pu) per gland to 3.5 x 10e10 pu per gland. Currently, we are pre-screening patients and determining enrollment eligibility according to approved inclusion and exclusion criteria.[unreadable] [unreadable] In collaboration with the Radiation Biology Branch of NCI, we continued to examine the usefulness of a non-gene therapy approach to prevent IR-induced salivary hypofunction in mice with the stable nitroxide Tempol. This year, we completed studies evaluating Tempol for differential radiation protection of salivary glands and tumor in a fractionated IR scheme. Tempol protected salivary glands against IR damage, without tumor protection. Intracellular reduction of Tempol to the non-protective hydroxylamine, assessed by MRI, was 2-fold faster in tumor versus gland or muscle. These results support further development of Tempol for human trials as a selective protector against IR-induced salivary gland damage.[unreadable] [unreadable] We also evaluated a preventive rAd5-mediated gene transfer approach for IR damage and tested the importance of murine gland endothelial cells to IR-induced injury. Specifically, we tested whether vector-mediated transfer of cDNAs for two angiogenic factors, basic fibroblast growth factor (AdbFGF) or vascular endothelial growth factor (AdVEGF), would afford glands radioprotection. If mice were pretreated with either vector 48 hours before IR microvessel loss was significantly reduced. After 8 weeks, IR reduced salivary flow 65% in untreated mice or mice treated with a control Ad5 vector. However, irradiated mice pretreated with AdbFGF or AdVEGF showed a significant improvement in their salivary flow, to 70% of unirradiated control mice. [unreadable] [unreadable] Our past studies in rodents showed that salivary glands are a potentially useful target for treating SSPDDs, particularly when the transgene encodes a constitutive pathway secretory protein secreted into the bloodstream. This year, we extended our studies in 2 large animal models, non-human primates and miniature pigs, a step critical in developing a gene therapy. We examined salivary gland administration of a serotype 2 adeno-associated viral (rAAV2) vector encoding rhesus erythropoietin (RhEpo) to rhesus parotid glands. The most important findings were (i) gene transfer had no negative effects on general macaque physiology (e.g., weight, complete blood count, serum chemistry), or gross and microscopic pathology; (ii) at necropsy, 6 months following administration, vector was overwhelmingly found in the targeted parotid gland; and (iii) the serum: saliva ratios of total RhEpo secreted showed that RhEpo secretion into the bloodstream was less than previously seen with human (h) Epo in male mice, i.e., serum: saliva ratio of 160:1, while in male macaques it was 7:1.[unreadable] [unreadable] We next tested in male macaques whether the vector used, or the encoded transgene, influenced the sorting observed. We used a rAd5 vector encoding hEpo, AdCMVhEpo, with 2 macaques per group. The dose administered (10e11 pu per gland) corresponded to a MOI of 2x10e7 pu per ul infused. We collected saliva and serum prior to, and on days 2, 7 and 14 after, vector administration. We did the same study in male Balbc mice. The serum: saliva ratios for total hEpo secreted were in excellent agreement with those found after rAAV2 vector-mediated gene transfer; in mice the ratio was 180:1, while in macaques it was 4.5:1, suggesting neither vector type, nor the specific Epo encoded (rhesus or human), were responsible for the observed species differences in Epo sorting. Also, since rAd5 vectors target acinar and duct cells, and rAAV2 vectors target only duct cells, it seems unlikely that the epithelial cell type targeted significantly influenced hEpo sorting.[unreadable] [unreadable] We also conducted similar experiments in male miniature pigs, examining hEpo secretion from parotid glands, using the same AdCMVhEpo at similar doses. hEpo was secreted into both saliva and serum, with most being found in saliva (>60-fold). However, serum hEpo levels were sufficient to significantly increase hematocrit levels by 10%. The amount of vector found in the targeted glands, at necropsy on day 14, was 100x more than in other tissues, i.e., similar to that in macaques with a rAAV2 vector. Overall, the results in these large animal model studies support our general hypothesis that salivary gland gene transfer can be used for treating SSPDDs, but also highlight significant species differences in secretory protein sorting.[unreadable] [unreadable] Additionally, we began to examine gender differences in salivary gland gene transfer with investigators at NIEHS. We conducted a large GLP safety study, testing 4 doses (10e7-10e10 pu) of rAAV2hEpo administered to 1 gland of male and female Balbc mice. Control and treated animals did not differ in clinical appearance, morbidity and mortality rates, food and water consumption, or weight, and no major vector related toxicity was seen on complete pathology and histopathology review. However, a significant gender related difference in vector distribution was revealed by qPCR. For example, at a dose of 10e10 pu per animal, vector effectively transduced, and was primarily confined within, male salivary glands (i.e., 800x more copies than liver; day 3) and was long lived. In contrast, in females salivary gland transduction was much less (260-fold < males; day 3) and short lived, and vector was disseminated widely into the bloodstream (salivary gland to liver copy ratio 1) and the saliva (30-fold > males). These studies suggest sexual dimorphism may be a factor of significance affecting clinical gene therapy applications in salivary glands.[unreadable] [unreadable] We have long worked on a novel secondary therapeutic approach; to develop an artificial salivary gland for patients unsuitable for direct gene therapy treatment. The envisioned device consists of a blind-end tube made of a biodegradable polymer (poly-L-lactate), coated on its luminal surface with fibronectin, upon which sits a monolayer of polarized epithelial cells capable of unidirectional fluid movement. We previously reported methods to obtain primary epithelial cells from human (huSMG) and rhesus (RPG) salivary glands capable of forming such a monolayer. This year we focused on engineering huSMG and RPG cells to secrete salt and water under physiological conditions, i.e., in which they would generate their own osmotic gradient, and shown this is possible.